Abstract

Parkinsons disease is a common neurodegenerative neurological disorder, which affects the patients quality of life, has significant social and economic effects, and is difficult to diagnose early due to the gradual appearance of symptoms. Examining the discussion of Parkinsons disease in social media platforms such as Twitter provides a platform where patients communicate each other in both diagnosis and treatment stage of the Parkinsons disease. The purpose of this work is to evaluate and compare the sentiment analysis of people about Parkinsons disease by using deep learning and word embedding models. To the best of our knowledge, this is the very first study to analyze Parkinsons disease from social media by using word embedding models and deep learning algorithms. In this study, Word2Vec, GloVe, and FastText are employed as word embedding models for the purpose of enriching tweets in terms of semantic, context, and syntax. Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Long Short-Term Memory Networks (LSTMs) are implemented for the classification task. This study demonstrates the efficiency of using word embedding models and deep learning algorithms to understand the needs of patients and provide a valuable contribution to the treatment process by analyzing sentiments of them with 93.63% accuracy performance.

References

[1] I. De Martino, R. D’Apolito, A.S. McLawhorn, K.A. Fehring, P.K. Sculco, G. Gasparini,Social media for patients: benefits and drawbacks, Curr. Rev. Musculoskelet. Med. 10 (2017)141–145.
[2] P. Eckler, G. Worsowicz, J.W. Rayburn, Social media and health care: an overview, Pm R. 2(2010) 1046–1050.
[3] V.M. Prieto, S. Matos, M. Álvarez, F. Cacheda, J.L. Oliveira, Twitter: a good place to detecthealth conditions, PloS One 9 (2014) e86191, , https://doi.org/10.1371/ journal.pone.0086191.
[4] L. Sinnenberg, A.M. Buttenheim, K. Padrez, C. Mancheno, L. Ungar, R.M. Merchant, Twitteras a tool for health research: a systematic review, Am. J. Public Health 107 (2017) e1–e8
[5] B.L. Neiger, R. Thackeray, S.H. Burton, C.R. Thackeray, J.H. Reese, Use of twitter amonglocal health departments: an analysis of information sharing, engagement, and action, J. Med.Internet Res. 15 (2013) e177, , https://doi.org/10.2196/jmir. 2775.
[6] C. Chew, G. Eysenbach, Pandemics in the age of twitter: content analysis of tweets during the2009 H1N1 outbreak, PloS One 5 (2010) e14118, , https://doi.org/10.1371/journal.pone.0014118.
[7] A. Beykikhoshk, O. Arandjelovic, D. Phung, S. Venkatesh, T. Caelli, Data-Mining Twitterand the Autism Spectrum Disorder: A Pilot Study, (2014), pp. 1–8.
[8] A. Beykikhoshk, O. Arandjelovic, D. Phung, S. Venkatesh, Overcoming Data Scarcity ofTwitter: Using Tweets as Bootstrap with Application to Autism-Related Topic Content Analysis,ACM, New York, New York, USA, 2015, https://doi.org/10. 1145/2808797.2808908.
[9] S. Fahn, Description of Parkinson's disease as a clinical syndrome, Ann N Y Acad Sci, 991(2003) 1-14.
[10] M.C. de Rijk, C. Tzourio, M.M. Breteler, J.F. Dartigues, L. Amaducci, S. Lopez-Pousa, J.M.Manubens-Bertran, A. Alperovitch, W.A. Rocca, Prevalence of parkinsonism and Parkinson'sdisease in Europe: the EUROPARKINSON Collaborative Study. European CommunityConcerted Action on the Epidemiology of Parkinson's disease, J Neurol Neurosurg Psychiatry, 62(1997) 10-15.
[11] Q. V. Le and T. Mikolov, “Distributed Representations of Sentences and Documents,” vol.32, 2014.
[12] Mikolov, T., Chen, K., Corrado, G., & Dean, J. (2013). Efficient estimation of wordrepresentations in vector space. arXiv preprint arXiv:1301.3781.
[13] J. Pennington, R. Socher, and C. Manning, “Glove: Global Vectors for WordRepresentation,” in Proceedings of the 2014 Conference on Empirical Methods in NaturalLanguage Processing (EMNLP), 2014.
[14] A. Joulin, E. Grave, P. Bojanowski, and T. Mikolov, “Bag of Tricks for Efficient TextClassification,” 2016.
[15] Kilimci, Z. H., & Akyokus, S. (2018). Deep Learning-and Word Embedding-BasedHeterogeneous Classifier Ensembles for Text Classification. Complexity, 2018.
[16] Ene, Marius. "Neural network-based approach to discriminate healthy people from thosewith Parkinson’s disease." Annals of the University of Craiova-Mathematics and Computer Science Series 35 (2008): 112-116.
[17] Little, M. A., McSharry, P. E., Hunter, E. J., Spielman, J., & Ramig, L. O. (2009).Suitability of dysphonia measurements for telemonitoring of Parkinson's disease. IEEEtransactions on bio-medical engineering, 56(4), 1015.
[18] Sakar, C. Okan, and Olcay Kursun. "Telediagnosis of Parkinson’s disease usingmeasurements of dysphonia." Journal of medical systems 34.4 (2010): 591-599.
[19] Das, Resul. "A comparison of multiple classification methods for diagnosis ofParkinson disease." Expert Systems with Applications 37.2 (2010): 1568-1572.
[20] Caglar, Mehmet Fatih, Bayram Cetisli, and Inayet Burcu Toprak. "AutomaticRecognition of Parkinson?s Disease from Sustained Phonation Tests Using ANN andAdaptive Neuro-Fuzzy Classifier." Journal of Engineering Science and Design 1.2 (2010):59-64.
[21] Polat, Kemal. "Classification of Parkinson's disease using feature weightingmethod on the basis of fuzzy C-means clustering." International Journal of SystemsScience 43.4 (2012): 597-609.
[22] Luukka, Pasi. "Feature selection using fuzzy entropy measures with similarityclassifier." Expert Systems with Applications 38.4 (2011): 4600-4607.
[23] Kihel, Badra Khellat, and Mohamed Benyettou. "Parkinson's Disease RecognitionUsing Artificial Immune System." JSEA 4.7 (2011): 391-395.
[24] Eskidere, Ö. (2012). A Comparison Of Feature Selection Methods For Diagnosis OfParkinson’s Disease From Vocal Measurements. Sigma, 30, 402-414.
[25] Prashanth, R., & Roy, S. D. (2018). Novel and improved stage estimation in Parkinson'sdisease using clinical scales and machine learning. Neurocomputing, 305, 78-103.
[26] Bruns, A., Kornstadt, A., & Wichmann, D. (2009). Web application tests with selenium.IEEE software, 26(5), 88-91.
[27] Loria, S. (2018). textblob Documentation (pp. 1-73). Technical report.
[28] Z. C. Lipton, J. Berkowitz, and C. Elkan, “A Critical Review of Recurrent Neural Networksfor Sequence Learning,” May 2015.
[29] J. L. Elman, “Finding structure in time,” Cogn. Sci., 1990.
[30] A. Graves, A. R. Mohamed, and G. Hinton, “Speech recognition with deep recurrent neuralnetworks,” in ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, 2013.
[31] Prashanth, R., & Roy, S. D. (2018). Early detection of Parkinson’s disease through patientquestionnaire and predictive modelling. International journal of medical informatics, 119, 75-87.
[32] J. Schmidhuber, “Deep Learning in neural networks: An overview,” Neural Networks, vol.61, pp. 85–117, 2015.
[33] G. H. Yann LeCun, Yoshua Bengio, “Deep learning (2015), Y. LeCun, Y. Bengio and G.Hinton,” Nature, 2015.
[34] A. Voulodimos, N. Doulamis, A. Doulamis, and E. Protopapadakis, “Deep Learning forComputer Vision: A Brief Review,” Comput. Intell. Neurosci., 2018.
[35] A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet Classification with DeepConvolutional Neural Networks,” in ImageNet Classification with Deep Convolutional NeuralNetworks, 2012.
[36] Zhang, L., Wang, S., & Liu, B. (2018). Deep learning for sentiment analysis: A survey.Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 8(4), e1253.
[37] Schneider KM. On Word Frequency Information and Negative Evidence in Naive BayesText Classification. In: 4th International Conference on Advances in Natural LanguageProcessing; 2004; Alacant, Spain: pp. 474–485.
[38] Kilimci, Z. H., Selim Akyokus, and Sevinc Ilhan Omurca. "The effectiveness ofhomogenous ensemble classifiers for Turkish and English texts." In INnovations in IntelligentSysTems and Applications (INISTA), 2016 International Symposium on, pp. 1-7. IEEE, 2016.
[39] T. Mikolov, K. Chen, G. Corrado, and J. Dean, “English pretrained Word2vec model,”https://code.google.com/archive/ p/word2vec/.